From: brian g. <bg...@us...> - 2003-01-13 19:02:13
|
Dave, wow, that's the kind of bug report i like. good eye, and thanks for the fix. also thanks for submitting the actual bug, cause it'll help me remember to include the fix on the next 1.3.x release. brian. On Sun, 12 Jan 2003, David J. Naffin wrote: > Brian > It would seem that attempting to connect to a player/stage simulation > with certain multiples of eight robots (i.e. 8/16/32 etc) > causes a segmentation violation. The problem is with the function > ClientManager::RemoveBlanks() in clientmanager.cc. It seems that the > last operation of recounting the number of connections has a flaw. There > is an assumption in the code that there is always a NULL client pointer > at the end of the list. This is true only when the list is not full. If > the list is full (i.e. num_clients = size_cleints) the very last for > loop in the function incorrectly counts the number of connections. To > fix this bug change line 295 in the file clientmanager.cc from: > > for(i=0;clients[i];i++); > > to > for(i=0;i<size_clients &&clients[i];i++); > > You are probably wondering "if he knows how to fixed it why does he just > submit the change himself?". Well I still have not figured out why I can > not commit changes on sourceforge. Since I have my CVS repository > working just the way I like it, I don't want to muck with the plumbing > right now. > > BTW I have submitted thais as an actual player bug as well. > > > Dave > > > > > > > > > ------------------------------------------------------- > This SF.NET email is sponsored by: > SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! > http://www.vasoftware.com > _______________________________________________ > Playerstage-developers mailing list > Pla...@li... > https://lists.sourceforge.net/lists/listinfo/playerstage-developers > > |